Elevated rates of DCIS diagnoses are inherent to current breast cancer screening processes - almost 30% of screen-detected breast cancers are DCIS. Due to uncertainty in the natural history of DCIS, there is widespread concern regarding overtreatment. Unfortunately, it is currently impossible to determine which DCIS lesions are likely to progress to a potentially lethal invasive stage. Thus, current guidelines recommend relatively aggressive treatment for all women with DCIS, including surgery, radiation, and consideration of hormone therapy. To optimize the breast cancer screening process, there is an urgent need for identification of DCIS prognostic markers that would permit personalized treatment strategies. Mammographic breast density is a promising candidate as a prognostic marker to predict the likelihood of progression from DCIS to invasive disease. Currently, however, there Is only scarce data regarding the nature of the association between breast density and disease progression, and much uncertainty in our understanding of the biological mechanisms of breast density. Collagen is a major component of breast density and laboratory studies have shown that it plays a key role in facilitating tumor Invasion. The objective of our proposal is to translate these laboratory findings into advances in the development of breast density as a prognostic marker for DCIS. We aim to 1) determine the association between mammographic breast density and disease-free survival among women with DCIS; 2) determine the association between collagen reorganization and disease-free survival among women with DCIS; and 3) assess whether the association between mammographic breast density and disease-free survival is mediated by collagen reorganization. To accomplish these aims, we will use data and tissue from the Vermont Breast Cancer Surveillance System, which includes linked patient risk factor, mammography, pathology, treatment, and cancer outcomes data for approximately 1,400 DCIS cases with up to 16 years of follow-up. We will use three different measures of breast density: the categorical BIF^DS assessment, a 2-D quantitative computer-assisted method (Cumulus), and a 3-D quantitative volumetric density assessment that permits measurement of breast density in specific regions of interest adjacent to the DCIS lesion. Multiphoton microscopy will be used to evaluate collagen reorganization in archived DCIS tumor specimens. This study will evaluate the potential for mammographic breast density and collagen reorganization to serve as potential markers for identifying DCIS cases that are not likely to progress or could be treated with only minimal intervention. This could lead to a substantial improvement in our ability to minimize the harms of breast cancer screening (overtreatment) while preserving the benefits (reductions in breast cancer morbidity and mortality).